Skip to content
Solopreneur · AI OS Operator

Scott MacLean · scottmaclean.org

One person.
A company of AI agents.

I run a 24/7 AI OS, a governed company of agents that takes an idea through research, build, deploy, market, and revenue.

I operate it solo and approve at the gates. The map beside this is that system, running. Every node is a real agent, colored by whether it's live, proven, in build, or held. This site is the proof.

Find me
01 / The system

The loop, named end to end

The AI OS is a governed venture loop. Each stage is a set of agents with a typed handoff to the next; autonomous between the gates, with the owner signing off at the gates. Each stage below carries its honest status.

01proven
Idea

A signed product blueprint enters and is checked.

02in build
Research

Scout verifies the perishable claims before build.

03held
Build

Production compiles the product, held on an app-builder.

04proven
Deploy

A versioned release, provisioned and smoke-tested.

05in build
Market

Go-to-market: the content engine proven, distribution building.

06in build
Revenue

Kaizen agents analyze the feedback and improve the next run.

Human gates between stages. The owner approves by design, not by exception.
02 / No overclaim

Where the system actually stands

The proof is Agent Lens, a real product the loop already shipped and put live. Everything below is the system behind it. The live builds themselves live where you can use them, on MacLean Market and my social channels.

Agent Lens
Live
The first product the loop shipped. It is live at agent-lens.com, grading how AI assistants read a website and returning the fix.
Planning
Proven in test
The compiler: blueprint → intake check → template → scope → sign → validate. Eight test blueprints run end to end, signed and verified.
Scout
In build
The research-verifier at the front of Planning. It takes the checkable claims a human produced and confirms they still hold, right now.
Production
Held
Builds on the UI kit and holds on a real capability gap: the first full vertical needs an app-builder. It reports the gap; it never fakes the build.
Deploy
Proven in test
Provision, release manifest, smoke test. Test products have gone live end to end on real infrastructure.
MacLean Media
In build
Go-to-market: the content engine is proven; distribution and press run owner-approved, and are being wired to the loop.
Kaizen
In build
The feedback loop: collect, cluster, score, and route signals into a delta the owner approves. Thin today, wiring to live signals.
03 / Proof the loop shipped

Agent Lens, the first live product

The loop is not a diagram. It has already produced a real, live product. Agent Lens grades how AI assistants read your website, and hands you the fix.

Live product

How AI assistants read your website: graded, with the fix.

As buyers ask AI assistants instead of searching, the question becomes: what does the AI actually see on your site? Agent Lens runs the scan, scores it, and returns the specific fixes.

Visit agent-lens.com
The engine

Systems thinking, made to ship

Typed contracts between stages. Deterministic checks where determinism belongs, judgment only where it earns its cost. Quality gates that hold instead of guessing.

See the loop
The storefront

Where the OS ships its products

MacLean Market is the front door for what the AI OS builds. Agent Lens is the first product through it; the loop is designed to send more.

Visit MacLean Market
04 / The operator

Who's behind the system

I'd rather build the machine that builds the product than build one product by hand.

What I'm really after is providing value, at scale. One person can only hand-build so much, so I've been building the system to do it instead: a one-person company of AI agents that does the work and can keep taking on more.

I build in the open and write the field notes as I go, on The MacLean Method.

RoleSolo operator, AI OS
First live productAgent Lens ↗

Two ways in

See what the AI OS ships, or read how it gets built. One system, one operator, in the open.